
@usekernel building the internet runtime for agents | prev @cashapp @afterpayUS
New Jersey, USA
Joined June 2016
- Tweets2.5K
- Following1.3K
- Followers806
- Likes8K
Pinned Tweet
which personal agent has the most magical cron-job management experience?
where i don’t have to always specify the frequency of checking on something. it just gets it consistently (>90%) without asking me to confirm.
this is easily a top 3 feature for me when choosing a managed personal agent.
big risk for personal agents: over-proactivity early on.
broad account perms (e.g. gmail, gcal) often let them act far beyond user expectations without asking. it's a fine line between creating a "wow" moment and overstepping. the real magic is knowing what actions not to take.
3
15
695
personal agents should be creatively using native polls in imessage for structured input instead of always writing walls of text
wired this up on my own agent using @thelinqapp poll api, so it can narrow down choices step by step. it’s fun to engage w/ a few of these thrown in
7
23
1,512
deeply optimistic for the agentic future, and glad to work somewhere whose vision aligns with that.
9
491
danny prevoznik retweeted
See the film: youtube.com/watch?v=deQsLCgN…
danny prevoznik retweeted
.@usekernel is rolling out our first ad ever and we hosted a party last night to celebrate.
We worked with an animation studio in France for 8 months to create it. Every scene was created by hand, down to the custom score.
You’ll see it around SF over the next several weeks: a mural on 10th and Howard, bus stops in FiDi, and @WorksInProgMag @arenamag ’s December issues.
We wanted to create something special to match the creativity and care we believe the future deserves.
Here's to the builders who are creating the future. built on KERNEL.
i embarked on a replit mission to make @usekernel documentation more fun to explore. the swarm (w/ a $30 budget it didn’t hit) read all the pages in our docs, build a 3d product map, and tested the output across device viewports.
link below if interested 👇
LLMs have gotten more autonomous recently but when you give them a goal they often misunderstand you, work aimlessly, and waste your tokens.
Today, we’re introducing Missions—the world’s first agentic swarm harness that can accomplish any goal you give them within set budget.
Try it today with something like “get me my first 10 customers:” replit.com/blog/introducing-…
you can walk through the fully interactive product experience yourself here:
kernel-product-map.replit.ap…
1
1
215
danny prevoznik retweeted
Today we’re announcing $90M in new funding, led by @a16z, and welcoming @bhorowitz to our board.
We've now raised over $120M to date.
We believe agents should be as easy to reach as a friend, and capable enough to finish the job. You should be able to text or call an agent and have it get real work done.
We started with messaging. Poke launched on Linq last year, and today @DoorDash, @TownAI, and @tomo run their agents on Linq. Our customers have now processed more than one billion messages.
DoorDash uses Linq to help customers find food, build a cart, and check out in a single thread.
And now, as an Apple Messaging Service Provider, we can help more enterprises launch and scale customer-facing agents on the same channel.
We expect conversational agents to reach every category of software.
We’ll ask them to find a place to live, coordinate with our teams, and manage our appointments. Businesses will build agents their customers can use to book, buy, and get things done.
Completing those requests often takes more than a conversation. An agent may need to call an office, use software on a computer, or come back to ask for approval.
That’s why Linq is expanding beyond messaging. Our new HD Calling product lets agents make handset-quality phone calls.
We’re also giving agents access to physical computing infrastructure so they can interact with software and complete tasks, with @usekernel among our first computer use customers.
We’ll use this funding to keep investing in messaging, expand our software and hardware infrastructure, and build the partnerships that make more powerful agents accessible to everyone.
An eternal thank you to our customers, our investors, and the team building it with us.
danny prevoznik retweeted
congrats to @elliott__potter and the Linq team! excited to share more soon about what we're cooking up with them at @usekernel
Today we’re announcing $90M in new funding, led by @a16z, and welcoming @bhorowitz to our board.
We've now raised over $120M to date.
We believe agents should be as easy to reach as a friend, and capable enough to finish the job. You should be able to text or call an agent and have it get real work done.
We started with messaging. Poke launched on Linq last year, and today @DoorDash, @TownAI, and @tomo run their agents on Linq. Our customers have now processed more than one billion messages.
DoorDash uses Linq to help customers find food, build a cart, and check out in a single thread.
And now, as an Apple Messaging Service Provider, we can help more enterprises launch and scale customer-facing agents on the same channel.
We expect conversational agents to reach every category of software.
We’ll ask them to find a place to live, coordinate with our teams, and manage our appointments. Businesses will build agents their customers can use to book, buy, and get things done.
Completing those requests often takes more than a conversation. An agent may need to call an office, use software on a computer, or come back to ask for approval.
That’s why Linq is expanding beyond messaging. Our new HD Calling product lets agents make handset-quality phone calls.
We’re also giving agents access to physical computing infrastructure so they can interact with software and complete tasks, with @usekernel among our first computer use customers.
We’ll use this funding to keep investing in messaging, expand our software and hardware infrastructure, and build the partnerships that make more powerful agents accessible to everyone.
An eternal thank you to our customers, our investors, and the team building it with us.
it came! thanks @iradotso for the assist and to @GenTentUSA for the t-shirt + great generator tent
couldn’t resist the siren call of a free t-shirt…an easy task for @iradotso going off just the photo taken
agentic commerce will happen slowly, then suddenly
Sweetgreen saw orders placed through AI agents double from June to October and are accelerating. Its still small but Sweetgreen sees more agent traffic than companies with many more times its volume.
(via @JonnyNemo on TBPN)
1
1
90
danny prevoznik retweeted
you dont work on bleeding edge stuff if your slack isnt littered with this guy



























